when the future starts today: infl ation expectations of ... · households form their infl ation...
TRANSCRIPT
93Annual Report 2018
Outlook and
Policy in 2019
77Annual Report 2018
When the Future Starts Today: Infl ation Expectations of Malaysian Households
By Zul-Fadzli Abu Bakar and Nur Aimi Abdul Ghani
Assessments on infl ation expectations are of great importance in central banking, both in advanced and emerging market economies. Central banks care about infl ation expectations given their prospective infl uence on key economic and fi nancial variables such as actual infl ation, households’ consumption and saving decisions, fi rms’ price- and wage-setting decisions, as well as fi nancial asset valuations.
The concept of infl ation expectations applies to households, fi rms, professional forecasters and fi nancial markets. While the beliefs of future infl ation hold across these different economic agents, their expectations are not interchangeable as they provide different insights (Table 1) (Coibion et al., 2018).
Given these different types of inflation expectations, this article focuses specifically on inflation expectations of households1 in Malaysia, using data from the Bank Negara Malaysia Consumer Sentiment Survey (BNM CSS). It aims to explain the formation of these expectations, seeking insights into how households process information to inform their future price outlook. It then explores factors that contribute to biases in inflation expectations. As inflation expectations may influence households’ spending decisions, the article also examines the relationship between inflation expectations and households’ spending decisions.
1 In the BNM CSS, a household is defi ned as related and/or unrelated persons who usually live together and make common provisions for food and other living essentials.
Table 1: Differences between Inflation Expectations of Households, Firms, Professional Forecasters and Financial Markets
Households’ view of how prices will evolve
in the future
Survey ofhouseholds
Consumption & saving decisions, wage
bargaining
Inflationary sentiments
of firms
Survey of firms
Pricing strategy, investment, hiring and wage-setting decisions
Expected inflation rate of professional economic
forecasters
Survey of professional economic forecasters
View on broader economic conditions
Market-implied, i.e. derived from prices of financial
instruments
Financial markets’ expectations of
future prices
Financial asset valuations
How are they commonly measured?
What insights dothey provide?
What are they?
Source: Coibion et al. (2018)
Households
Financialmarkets
Professionalforecasters
Firms
94 Annual Report 2018
Outlook and
Policy in 2019
78 Annual Report 2018
Infl ation Expectations, Infl ation Dynamics and the Conduct of Central BankingOver the years, there has been a marked increase in the number of central banks that factor infl ation expectations into their monetary policy considerations (Diagram 1). This trend is largely grounded upon economic theory which posits that infl ation expectations are salient in infl uencing the behaviour of economic agents, making them a key driving force of actual infl ation. The New Keynesian Phillips Curve (NKPC), which is a theoretical framework for infl ation dynamics, explicitly incorporates this forward-looking expectations component. A property of the NKPC is that infl ation is a forward-looking process, driven by expectations of future infl ation and not merely past and current shocks.
1. Why are infl ation expectations important?
Often, we may hear statements like these from friends, families, or strangers in our daily lives. In essence, these statements are based on what is referred to as inflation expectations: the belief of how much prices will change in the future. In the modern economy, where every household is affected by price changes, it is reasonable to infer that each decision-maker holds some belief about future inflation (Mohanty, 2012).
Understanding inflation expectations is important because they affect the real economy in a number of ways (Moore, 2016). First, for firms, inflation expectations affect price-setting decisions. If firms expect higher inflation, they may be more inclined to set higher prices, which would lead to higher actual inflation. Second, in countries with strong labour bargaining power, inflation expectations affect wage negotiations. If inflation is expected to be persistently higher, workers bargain for higher wages, which places upward pressure on firms to increase prices. Third, for households in general, inflation expectations affect current spending decisions.
Diagram 1: Statements on Inflation Expectations among Policymakers Across the Globe
“Inflation expectations are a key determinantof actual inflation and are thus a crucial part ofthe analysis used by many central banks togenerate inflation forecasts.” – BOC, 2010
“In our thinking, inflation expectationsare the most important driver of actualinflation.” – the Fed, 2019
“…inflation expectations are of great importance for the conduct of monetary policy and are closely monitored by the ECB.” – ECB, 2011
“...MAS monitors developments in wages and rentals, … and indicators of inflation expectations as these contain information about future price dynamics.” – MAS, 2018
“Increased inflation expectations, risk of second-hand effects call for monetary policy response.” – BSP, 2018
“Inflation expectations are important for a number of reasons.” – RBA, 2016
Source: The Federal Reserve System (the Fed), Bank of Canada (BOC), European Central Bank (ECB), Banco Central de la República Argentina (BCRA), Bangko Sentral ng Pilipinas (BSP), Reserve Bank of Australia (RBA) and Monetary Authority of Singapore (MAS)
“Once a Central Bank has gained credibility for the achievements of its long-term targets, it will be able to align agents’ expectations with future inflation targets.”- BCRA, 2016
“Looks like prices are going to go up by quite a lot… I’d better buy this TV now before it gets even more expensive!”
“Yes you’re right, but I think I’ll start saving up in case there is an emergency in the future”
“Ya betul, tapi saya rasa saya akan mula berjimat untuk kecemasan pada masa depan”
“Nampaknya harga akan semakin meningkat dengan lebih banyak... Baik saya beli TV ini sekarang sebelum harga jadi lebih mahal!”
95Annual Report 2018
Outlook and
Policy in 2019
79Annual Report 2018
2. How are household infl ation expectations measured?
Household infl ation expectations are measured through surveys
Unlike actual infl ation itself, infl ation expectations are unobserved. As such, their measurements rely on inference. A standard approach globally is to directly question households what they expect infl ation to be in the following year through periodic surveys catered to extract household economic information (Table 2). There are variations to this question, which may be in qualitative or quantitative form, and either general or specifi c:2
For Malaysia, similar questions are featured as part of the BNM CSS. Each month, around 1,000 respondents are interviewed from samples designed to be representative of the Malaysian population, with respondents ranging from 18 to 74 years old. The BNM CSS commenced in 2013 to bridge information gaps pertaining to consumer expectations on economic growth, infl ation, wage growth and borrowing cost.3 This survey also gathers households’ demographics such as their household size, income and location.
Infl ation expectations affect actual infl ation because they infl uence spending, price- and wage-setting decisions. They can cause a long-lasting impact to actual infl ation, whereby one round of price increases triggers further rounds as infl ationary psychology takes hold. Consequently, modern central banking practices involve anticipating future infl ation and managing infl ation expectations to ensure price stability. From a policy perspective, this forward-looking element opens the door to expectations management and communications as added tools of monetary policy.
Notwithstanding this, the sources of infl ation expectations, as well as the actual drivers of infl ation itself, need to be assessed holistically for monetary policy considerations. Notably, whether infl ation expectations and actual infl ation are demand- or supply-driven. The early 2007 – July 2008 period provides a good example, a time when global commodity prices, as measured by the IMF Commodity Price Index, increased by 83%, leading to rising global infl ation. In the euro area, concerns of rising infl ation expectations were at the forefront of policy discussions as there were risks of second-round effects to actual prices from price- and wage-setting behaviour. In other words, if rising infl ation expectations were not managed, the commodity price shock may have had longer-lasting effects to actual infl ation if fi rms were setting higher prices and households were bargaining for higher wages in anticipation of the higher future infl ation. This, along with other considerations, led the European Central Bank (ECB) to increase its policy rate by 25 basis points in July 2008.
In Malaysia, the monetary policy stance was maintained during this period despite the increase in supply-driven infl ation pressures. While the risks to infl ation were on the upside, it was assessed that the risks to growth were on the downside given the weakening global growth, and thus reduced the likelihood of second-round effects to infl ation. As the global fi nancial crisis intensifi ed, policy rates were consequently reduced by 150 basis points from October 2008 to February 2009.
Inflation expectations are only one of the many determinants of inflation. Ultimately, monetary policy must give due consideration to the overall inflation outlook, in addition to economic growth prospects and risks of financial imbalances.
2 These questions are adapted from the University of Michigan Survey of Consumers for American households (Meyer and Venkatu, 2011).3 The Malaysian Institute of Economic Research (MIER) also conducts a similar survey on consumer sentiments in Malaysia. The main
differences between MIER’s survey and that of BNM's are the coverage of topics and frequency of survey.
Qualitative and general: “During the next 12 months, do you think that prices in general will go up, go down, or remain the same?”
Quantitative and specifi c: “During the next 12 months, how much do you think prices of fuel will go up, go down, or do you think they will remain the same?”
96 Annual Report 2018
Outlook and
Policy in 2019
80 Annual Report 2018
Table 2: Surveys on Infl ation Expectations of Households in Various Countries
Country Survey Year started Frequency Coverage Expectation horizon
United StatesUniversity of Michigan, Survey of Consumers
1946 MonthlyMinimum of 500
households 1-year & 5-year ahead
EUEuropean Commission, Business and Consumer
Survey1961 Monthly
More than 41,000 households
1-year ahead
Australia Melbourne Institute, Consumer Attitudes,
Sentiments and Expectations in Australia Survey
1974 Monthly 1,200 households 1-year ahead
JapanBank of Japan, Opinion Survey on the General
Public’s Views and Behavior1993 Quarterly 4,000 households 1-year & 5-year ahead
New ZealandReserve Bank of New
Zealand, Household Infl ation Expectations Survey
1995 Quarterly 1,000 households 1-year & 5-year ahead
CanadaBank of Canada, the Canadian Survey of
Consumer Expectations2015 Quarterly 1,000 households
1-year, 2-year & 5-year ahead
IndonesiaBank Indonesia,
Consumer Expectation Survey
1999 Monthly 4,600 households 6-month ahead
IndiaReserve Bank of India,
Infl ation Expectations Survey of Households
2005 Quarterly 6,000 households 3-month & 1-year ahead
SingaporeSKBI-MasterCard Index of Infl ation
Expectations Survey2011 Quarterly 400 households 1-year & 5-year ahead
MalaysiaBNM Consumer
Sentiment Survey2013 Monthly 1,000 households
1-year & 2- to 3-year ahead
Source: University of Michigan, European Commission, Melbourne Institute, Bank of Japan, Reserve Bank of New Zealand, Bank of Canada, Bank Indonesia, Reserve Bank of India, Singapore Management University, and Bank Negara Malaysia
3. How do Malaysian households form their infl ation expectations?
Households form their infl ation expectations through a combination of past, current and forward-looking beliefs
To interpret the level and changes of infl ation expectations, it is useful to understand how they are formed. Households place different weights on a host of factors when constructing their infl ation expectations (Rowe, 2016). On the one hand, “backward-looking” households develop infl ation expectations by extrapolating past and current infl ation experiences into the future. These past and current infl ation experiences are also referred to as infl ation perceptions, which are the beliefs of how much prices have changed (Table 3). Sentiments surrounding income also shape infl ation perceptions because changes in income relative to price affect the perceived impact of price changes, and thereby households’ views toward affordability (Ranyard et al., 2008; Gärling and Gamble, 2006; Gamble, 2006).
81Annual Report 2018
On the other hand, “forward-looking” households consider expectations of future broader economic developments, such as their sentiments on the economy. In reality, households can integrate both backward- and forward-looking factors in shaping their infl ation expectations.
On average, Malaysian households place weights on both backward- and forward-looking factors (Chart 1).4 Past sentiments on the domestic economy, recent shopping experiences and current income levels are important backward-looking factors, reflecting their role as easy, accessible and personal ways to form predictions. The forward-looking aspect is also important, with households also considering their expectations of the future strength of the Malaysian economy when forming inflation expectations.
4 The estimates are derived using an Ordinary Least Squares (OLS) regression of infl ation expectations against factors identifi ed in the literature to affect infl ation expectations. The regression also controls for time effects. The estimates shown are statistically signifi cant at the 1% level.
Table 3: Differences Between Infl ation Expectations, Infl ation Perceptions and Actual Infl ation
Infl ation expectations Infl ation perceptions Actual infl ation
Defi nitionThe belief of how much prices will
change in the future
The belief of howmuch prices have
changed from the past
How much prices have changed from last year, based on a single basket of goods and services in the Consumer Price Index (CPI)
Infl uenced byIndividual’s information
sets of the past, present and future
Individual’s information sets of the
past and present
Average consumption patterns of Malaysian households and
average prices in the economy
Source: Bank Negara Malaysia, Arioli et al. (2017)
Chart 1: Selected Factors Estimated to Determine Inflation Expectations of Malaysian Households
Households consider both backward- and forward-looking factors when forming inflation expectations
Note: Other factors that are considered include expectations for future interest rate on savings, expectations for future interest rates on loans and various sources of inflation information such as the mass media and internet. The factors chosen are based on Rowe (2016).
0 5 10 15 20
Expectations for future strengthof the Malaysian economy
Household income
Actual inflation1
Recent shopping experiences
Past sentiments about theMalaysian economy
Past and current
Future
Relative importance (%)2
1 Actual inflation refers to Consumer Price Index (CPI) inflation of the current and past three months2 Relative importance refers to how much the factors contribute to the movements in inflation expectations relative to other factors considered
Source: Bank Negara Malaysia estimates using data from the Bank Negara Malaysia Consumer Sentiment Survey and Department of Statistics, Malaysia
97Annual Report 2018
Outlook and
Policy in 2019
81Annual Report 2018
On the other hand, “forward-looking” households consider expectations of future broader economic developments, such as their sentiments on the economy. In reality, households can integrate both backward- and forward-looking factors in shaping their infl ation expectations.
On average, Malaysian households place weights on both backward- and forward-looking factors (Chart 1).4 Past sentiments on the domestic economy, recent shopping experiences and current income levels are important backward-looking factors, reflecting their role as easy, accessible and personal ways to form predictions. The forward-looking aspect is also important, with households also considering their expectations of the future strength of the Malaysian economy when forming inflation expectations.
4 The estimates are derived using an Ordinary Least Squares (OLS) regression of infl ation expectations against factors identifi ed in the literature to affect infl ation expectations. The regression also controls for time effects. The estimates shown are statistically signifi cant at the 1% level.
Table 3: Differences Between Infl ation Expectations, Infl ation Perceptions and Actual Infl ation
Infl ation expectations Infl ation perceptions Actual infl ation
Defi nitionThe belief of how much prices will
change in the future
The belief of howmuch prices have
changed from the past
How much prices have changed from last year, based on a single basket of goods and services in the Consumer Price Index (CPI)
Infl uenced byIndividual’s information
sets of the past, present and future
Individual’s information sets of the
past and present
Average consumption patterns of Malaysian households and
average prices in the economy
Source: Bank Negara Malaysia, Arioli et al. (2017)
Chart 1: Selected Factors Estimated to Determine Inflation Expectations of Malaysian Households
Households consider both backward- and forward-looking factors when forming inflation expectations
Note: Other factors that are considered include expectations for future interest rate on savings, expectations for future interest rates on loans and various sources of inflation information such as the mass media and internet. The factors chosen are based on Rowe (2016).
0 5 10 15 20
Expectations for future strengthof the Malaysian economy
Household income
Actual inflation1
Recent shopping experiences
Past sentiments about theMalaysian economy
Past and current
Future
Relative importance (%)2
1 Actual inflation refers to Consumer Price Index (CPI) inflation of the current and past three months2 Relative importance refers to how much the factors contribute to the movements in inflation expectations relative to other factors considered
Source: Bank Negara Malaysia estimates using data from the Bank Negara Malaysia Consumer Sentiment Survey and Department of Statistics, Malaysia
98 Annual Report 2018
Outlook and
Policy in 2019
82 Annual Report 2018
Households who are pessimistic on economic or fi nancial conditions may be more likely to have higher infl ation expectations. As these households struggle to make ends meet, either due to insuffi cient income or rising prices, they are likely to anticipate higher future infl ation (Ehrmann et al., 2015). In other countries, it was found that fi nancially constrained households also pay more attention to price changes compared to less constrained households, and that households in general are more receptive to bad news than good news. In totality, this could lead to higher infl ation expectations (Ehrmann et al., 2015; Snir and Levy, 2011, Baumeister et al., 2001).
Most Malaysian households’ infl ation expectations are anchored at a relatively stable rate of 1 – 4%, though disparity is still observed
Some households tend to over-predict future infl ation, which is also referred to as positive expectations bias, while others may be more inclined to under-predict, which is called negative infl ation expectations bias. This disparity is evident in Malaysia when the infl ation expectations bias is grouped by demographic groups (Chart 2). Lower-income households, large households, working age individuals, households with less job security, and households in Kuala Lumpur have greater tendency for higher infl ation expectations.5 Some of the groups, such as lower-income households, consume more necessities (e.g. food) which typically exhibit higher rates of infl ation. The tendency for upward bias for these groups may also refl ect their cost of living burden and concerns about insuffi cient income. Indeed, a common grievance voiced by these groups is that the actual infl ation rate is not refl ective of the actual price increases they experience on the ground.
5 The estimates are derived using an Ordinary Least Squares (OLS) regression of infl ation expectations bias against demographic characteristics to quantify their contributions to the over-prediction or under-prediction of future infl ation. All the estimates shown are statistically signifi cant at the 10% level, except for elderly respondents and respondents residing in less-urbanised states.
Chart 2: Contribution of Main Demographic Characteristics to Inflation Expectations Bias
Certain households have greater tendency towards higher inflation expectations
Relative inflation expectations bias1 (percentage points)
-1.0
-0.5
0.0
0.5
1.0
1.5
Low
Hig
h
Sm
all
Larg
e
Wor
king
age
Eld
erly
Full
time
Tem
por
ary
Kua
la L
ump
ur
Income level Householdsize
Age Employmentstatus
Regions
Less
-urb
anis
ed s
tate
s2
Note: For income level, low refers to monthly household income below RM5,000 while high, above RM8,000. For household (HH) size, small refers to number of HH members below four while large, above seven. For age, working age refers to respondents aged between 20-59 while elderly, 60-74. For employment status, full time refers to permanent employees working eight hours a day or five days a week, while temporary refers to employees under a short-term, non-contractual or ad-hoc work arrangement. The methodology is based on Leung (2009)
1 The bias refers to 1-year ahead inflation expectations minus actual inflation outcomes. The relative inflation expectations bias should be interpreted relative to a baseline respondent who represents a HH that earns RM5,000-RM8,000, has a HH size of between 4-7 members, is aged between 18-19, is self-employed (e.g. sole proprietor or independent professional) and is residing outside Kuala Lumpur and less-urbanised states2 Less-urbanised states are states with urbanisation levels of below 55%, according to the 2010 Population and Housing Census of Malaysia. These states include Sabah, Sarawak, Pahang and Kelantan
Source: Bank Negara Malaysia estimates using data from the Bank Negara Malaysia Consumer Sentiment Survey and Department of Statistics, Malaysia
83Annual Report 2018
Notwithstanding the disparity across groups, for most periods, overall households’ infl ation expectations deviate on the upside from the actual infl ation outcomes (Chart 3). This observation is not unique to Malaysia. Households, in both advanced and emerging economies, commonly display deviations in infl ation expectations from the actual infl ation outcomes (Coibion et al. 2018, Kliesen, 2015, Mohanty, 2012, Gnan et al. 2009, Deutsche Bundesbank, 2001). For countries with a longer history of household survey data, namely the United States, Australia, Japan and the Philippines, the positive bias can range between 1 and 4 percentage points. On average, most households’ infl ation expectations in Malaysia are broadly anchored at a relatively stable rate of 1 – 4%, close to the actual infl ation long-run average of 3% (1980 – 2018).
Chart 3: Inflation Expectations of Malaysian Households
Actual InflationInflation Expectations (as reported one year ago)
-2
0
2
4
6
8
A M J J A S O N D J F M A M J J A S O N D J F M A M J J A S O N D J F M A M J J A S O N D
Positive bias
2018 201920172016
Negative bias
Annual growth (%)
In line with experiences of other countries, inflation expectations of Malaysian households are typicallydifferent from the actual inflation outcomes
Note: The shaded area refers to the 1 standard deviation above and below the mean inflation expectations
Source: Bank Negara Malaysia estimates using data from the Bank Negara Malaysia Consumer Sentiment Survey and Department of Statistics, Malaysia
Average Household Inflation Expectations Bias Across Selected Countries Since 2015
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
Japan Australia New Zealand Malaysia Philippines United States2
Inflation expectations bias1 (percentage points)
United Kingdom2
1 The bias is calculated by taking 1-year ahead mean inflation expectations minus actual inflation outcomes2 Inflation expectations for the United States and the United Kingdom are based on the median instead of the mean
Source: Bank Negara Malaysia estimates using data from the Bank Negara Malaysia Consumer Sentiment Survey, Department of Statistics, Malaysia and the respective household surveys
99Annual Report 2018
Outlook and
Policy in 2019
83Annual Report 2018
Notwithstanding the disparity across groups, for most periods, overall households’ infl ation expectations deviate on the upside from the actual infl ation outcomes (Chart 3). This observation is not unique to Malaysia. Households, in both advanced and emerging economies, commonly display deviations in infl ation expectations from the actual infl ation outcomes (Coibion et al. 2018, Kliesen, 2015, Mohanty, 2012, Gnan et al. 2009, Deutsche Bundesbank, 2001). For countries with a longer history of household survey data, namely the United States, Australia, Japan and the Philippines, the positive bias can range between 1 and 4 percentage points. On average, most households’ infl ation expectations in Malaysia are broadly anchored at a relatively stable rate of 1 – 4%, close to the actual infl ation long-run average of 3% (1980 – 2018).
Chart 3: Inflation Expectations of Malaysian Households
Actual InflationInflation Expectations (as reported one year ago)
-2
0
2
4
6
8
A M J J A S O N D J F M A M J J A S O N D J F M A M J J A S O N D J F M A M J J A S O N D
Positive bias
2018 201920172016
Negative bias
Annual growth (%)
In line with experiences of other countries, inflation expectations of Malaysian households are typicallydifferent from the actual inflation outcomes
Note: The shaded area refers to the 1 standard deviation above and below the mean inflation expectations
Source: Bank Negara Malaysia estimates using data from the Bank Negara Malaysia Consumer Sentiment Survey and Department of Statistics, Malaysia
Average Household Inflation Expectations Bias Across Selected Countries Since 2015
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
Japan Australia New Zealand Malaysia Philippines United States2
Inflation expectations bias1 (percentage points)
United Kingdom2
1 The bias is calculated by taking 1-year ahead mean inflation expectations minus actual inflation outcomes2 Inflation expectations for the United States and the United Kingdom are based on the median instead of the mean
Source: Bank Negara Malaysia estimates using data from the Bank Negara Malaysia Consumer Sentiment Survey, Department of Statistics, Malaysia and the respective household surveys
100 Annual Report 2018
Outlook and
Policy in 2019
84 Annual Report 2018
Source: Bank Negara Malaysia Consumer Sentiment Survey
Diagram 2: Household Sources of Information about Inflation
2015 2018
Family orfriends24%
Internet9%
Shoppingexperience
20%
Mass media 45%
Academicarticles
1%
29%Mass media
26%
Family orfriends
24%
Shoppingexperience
2%
Academicarticles
17%Internet
The role of the internet as a source of infomation on inflation has markedly increased over time
4. Why do infl ation expectations differ from actual infl ation?
Infl ation expectations are unlikely to be the same as the actual infl ation outcomes because forming accurate beliefs about future prices is challenging and costly, even for professional forecasters. It is found that in countries with extremely high infl ation such as Argentina, households appear to be much more well-informed on infl ation as its impact is acutely felt; whereas in countries with low infl ation, households pay less attention to infl ation given the relatively lower potential fi nancial cost of ignoring infl ation (Cavallo et al., 2017). In Malaysia, when respondents were asked whether they understood what the term “infl ation” means, 77% of them answered no. The low awareness among households would likely contribute to biases in their infl ation expectations. The wedge between infl ation expectations and actual infl ation outturns is driven by diverse information sets, different spending patterns and unanticipated shocks.
First, diverse information sets. Households rely on various information sources in informing their views on price developments. In Malaysia, households appear to be largely dependent on information sources that are relatively accessible when forming infl ation expectations. These include mass media, family and friends, as well as recent shopping experiences (Diagram 2).
These information sets and their importance are different from those used to construct the actual infl ation rate, which is measured by the Consumer Price Index. This could lead to infl ation expectations being different from the actual infl ation outcome.
Second, the different types of items consumed. Households typically over-estimate future infl ation of frequently-purchased items and necessities, such as transport and food (Diagram 3). Repeated shopping experiences of the same items (usually non-durable necessities), coupled with their corresponding price increases are more easily recalled and are therefore vivid in one’s recollection when forming expectations. It was estimated that households in Malaysia who rely on recent shopping experiences as an information source tend to over-predict infl ation by 1.2 percentage points (ppt) compared to households who do not rely on recent shopping experiences. This is compounded by memory bias, whereby increases in prices are more easily recalled rather than price declines (Loke and Khong, 2017).
101Annual Report 2018
Outlook and
Policy in 2019
85Annual Report 2018
-0.5 0 0.5 1
Clothing
Bills/ Utility
Electronics
Food away from home
Food at home
Transport
Less frequently-purchased: +0.1 ppt
Frequently-purchased: +1.7 percentage points (ppt)
Source: Bank Negara Malaysia estimates using data from the Bank Negara Malaysia Consumer Sentiment Survey and Department of Statistics, Malaysia
Diagram 3: Contribution of Key Consumer Items to Inflation Expectations Bias
Households have a higher tendency to over-predict the future inflation of frequently-purchased and necessity items
Contribution to inflation expectations bias
Necessities Discretionary Items
1.1 ppt 0.7 ppt
Contribution to inflation expectations bias
Third, unanticipated shocks. Even if households are able to use all existing information to inform future infl ation expectations, there remains a possibility of unforeseen events. In 2018, the actual infl ation outcome was lower than expected mainly due to the zerorisation of the Goods and Services Tax (GST) rate,6 which led to a broad-based decline in the prices of goods and services that were previously subjected to the GST. As a result, there was a positive bias as infl ation expectations of households over-predicted actual infl ation outcomes by an average of 3.4 ppt (Chart 4).
6 The zerorisation of the GST rate was between 1 June 2018 and 31 August 2018. Subsequently, the GST was replaced by the Sales and Services Tax (SST) effective 1 September 2018.
Source: Bank Negara Malaysia estimates using data from the Bank Negara Malaysia Consumer Sentiment Survey and Department of Statistics, Malaysia
-2
-1
0
1
2
3
4
5
N D J F M A M J J A S O N D J F M A M J J A S O N D J F M A M J J A S O N D
2016 2017 20182015
Over-prediction
Inflation expectations bias*(percentage points)
Under-prediction
*The bias is calculated by taking 1-year ahead inflation expectations minus actual inflation outcomes
Inflation expectations bias is subject to unanticipated shocks to the actual inflation outcomes
GSTzerorisation
OPECproduction cut
2016: 1.6 ppt2017: -0.3 ppt2018: 3.4 ppt
Chart 4: Inflation Expectations Bias: Over- and Under-Predictions Relative to Actual Inflation
102 Annual Report 2018
Outlook and
Policy in 2019
86 Annual Report 2018
In 2017, the actual inflation outcome was higher than anticipated due to global oil prices that increased unexpectedly following the Organization of the Petroleum Exporting Countries (OPEC)’s agreement to cut production from 1 January 2017.7 Subsequently, the inflation expectations of households under-predicted actual inflation, with a negative bias of -0.3 ppt.
5. Do infl ation expectations of households affect spending decisions?
Higher infl ation expectations could increase or decrease households’ current spending (Table 4). Higher infl ation expectations would lead to higher current real spending when households decide to frontload future purchases to protect themselves against a costlier future environment. In addition, assuming unchanged nominal interest rates, higher infl ation expectations would lower real interest rates, thereby encouraging borrowers to spend.
Higher current real spending from higher infl ation expectations:
Type ofhouseholds Reasons
General
Frontloading of future purchases
Higher infl ation expectations signal that it is costlier to spend in the future, so it is favourable to frontload future spending today
Borrowers
Lower real interest rates for debt repayments
The lower real interest rates lead to a perceived increase in wealth for borrowers. If borrowers have higher propensity to consume out of wealth than lenders, aggregate spending will increase
Lower current real spending from higher infl ation expectations:
Type of households Reasons
General
Uncertainty and increased precaution
Higher infl ation expectations may signal an environment of economic uncertainty, whereby households will tend to withhold spending and increase savings for precautionary purposes
Financialinvestors
Lower real income from fi nancial assets
Higher infl ation expectations are a tax on real fi nancial income due to the lower expected real returns. For households that rely on fi nancial income, this discourages spending
Source: Mian et al. (2013), Bachmann et al. (2015) and Coibion et al. (2018)
Table 4: The Impact of Higher Infl ation Expectations to Current Real Spending
7 The stronger-than-expected global oil demand since the second quarter of 2017 amid production cuts led to falling inventories, which also supported the increase in oil prices. The rise in global oil prices was further compounded by unexpected supply disruptions in the US Gulf Coast due to Hurricane Harvey at the end of August and geopolitical tensions in the Middle East towards the end of the year.
On the other hand, higher infl ation expectations could lead to a decline in current real spending if households associate the higher infl ation expectations with broader economic uncertainty, thereby preferring to save for precautionary reasons. For fi nancial investors, the lower real interest rates also reduce expected real income from fi nancial assets, discouraging spending. Overall, the effect of infl ation expectations on real spending would depend on which of these factors dominate.
Households' sentiments on current prices, income, the overall economy, and their debt burden also infl uence their spending decisions (Diagram 4). For Malaysian households, while sentiments on income and current prices are the major factors that infl uence reported current real spending, there is still a role for infl ation expectations, though small. Notably, there are other more prominent determinants of real spending decisions in Malaysia that are not captured by the BNM CSS, such as wealth (Bank Negara Malaysia, 2013).
87Annual Report 2018
* Relative importance refers to how much the sentiments contribute to reported current real spending, relative to the other sentiments considered
Source: Bank Negara Malaysia estimates using data from the Bank Negara Malaysia Consumer Sentiment Survey
Households consider, among others, their sentiments on current prices and income when makingspending decisions
Diagram 4: Relative Importance* of Various Types of Sentiments to Reported Current Real Spending
Sentiments on Economic Conditions
10%
Sentiments on Current Price Changes
57%
Sentiments on Future Price Changes
9%Sentiments on Debt4%
Sentiments on Income
20%
Note: The methodology is based on Ichiue and Nishiguchi (2013)
6. Conclusion and policy discussions
Infl ation expectations matter because they affect key economic and fi nancial variables such as actual infl ation, wage, spending and fi nancial asset valuations. Looking specifi cally at infl ation expectations of households in Malaysia, households form their infl ation expectations by considering both backward- and forward-looking factors, such as recent shopping experiences and current income levels, as well as the future outlook of the domestic economy. Nevertheless, infl ation expectations differ across households, given the different compositions, priorities and experiences. Certain demographic groups that typically experience higher cost of living burden, such as lower-income and large households, tend to display higher infl ation expectations.
As commonly experienced by other countries, there is evidence of a wedge between households’ infl ation expectations and actual infl ation outturns. This refl ects households’ broad inattention to infl ation, varied information sets used to gain information about infl ation, different consumption patterns, as well as unanticipated shocks to the economy.
However, the impact of households’ infl ation expectations on their current spending decisions is small, refl ecting other more prominent determinants such as sentiments on income and current prices in driving consumption. The implication of this fi nding, therefore, needs to be considered with care and in the broader context of factors driving household consumption spending. Separately from consumption, higher infl ation expectations could result in higher actual future infl ation if they become entrenched, whereby one round of price increase triggers further rounds as infl ationary psychology takes hold. As such, it is important for central banks to place focus on ensuring that these expectations are stable and anchored, rather than engineering higher expectations to meet short-term economic goals. This goes back to the need for having a solid understanding of how households form infl ation expectations, in order for any expectation anchoring strategies to be effective.
103Annual Report 2018
Outlook and
Policy in 2019
87Annual Report 2018
* Relative importance refers to how much the sentiments contribute to reported current real spending, relative to the other sentiments considered
Source: Bank Negara Malaysia estimates using data from the Bank Negara Malaysia Consumer Sentiment Survey
Households consider, among others, their sentiments on current prices and income when makingspending decisions
Diagram 4: Relative Importance* of Various Types of Sentiments to Reported Current Real Spending
Sentiments on Economic Conditions
10%
Sentiments on Current Price Changes
57%
Sentiments on Future Price Changes
9%Sentiments on Debt4%
Sentiments on Income
20%
Note: The methodology is based on Ichiue and Nishiguchi (2013)
6. Conclusion and policy discussions
Infl ation expectations matter because they affect key economic and fi nancial variables such as actual infl ation, wage, spending and fi nancial asset valuations. Looking specifi cally at infl ation expectations of households in Malaysia, households form their infl ation expectations by considering both backward- and forward-looking factors, such as recent shopping experiences and current income levels, as well as the future outlook of the domestic economy. Nevertheless, infl ation expectations differ across households, given the different compositions, priorities and experiences. Certain demographic groups that typically experience higher cost of living burden, such as lower-income and large households, tend to display higher infl ation expectations.
As commonly experienced by other countries, there is evidence of a wedge between households’ infl ation expectations and actual infl ation outturns. This refl ects households’ broad inattention to infl ation, varied information sets used to gain information about infl ation, different consumption patterns, as well as unanticipated shocks to the economy.
However, the impact of households’ infl ation expectations on their current spending decisions is small, refl ecting other more prominent determinants such as sentiments on income and current prices in driving consumption. The implication of this fi nding, therefore, needs to be considered with care and in the broader context of factors driving household consumption spending. Separately from consumption, higher infl ation expectations could result in higher actual future infl ation if they become entrenched, whereby one round of price increase triggers further rounds as infl ationary psychology takes hold. As such, it is important for central banks to place focus on ensuring that these expectations are stable and anchored, rather than engineering higher expectations to meet short-term economic goals. This goes back to the need for having a solid understanding of how households form infl ation expectations, in order for any expectation anchoring strategies to be effective.
104 Annual Report 2018
Outlook and
Policy in 2019
88 Annual Report 2018
To gain a complete understanding of the impact of infl ation expectations on the Malaysian economy, infl ation expectations of fi rms and fi nancial markets should also be assessed in order to obtain a comprehensive picture that includes insights on price- and wage-setting decisions, as well as fi nancial market valuations.
As ruminated by Tan Sri Dato’ Abdul Aziz bin Haji Taha, the second Malaysian Governor of Bank Negara Malaysia, nearly four decades ago – “Once businessmen and consumers discern an absence of will by the authorities to resist infl ation, their expectations of rising infl ation will be strengthened and they begin to act in a way that will make these expectations self-fulfi lling” (Bank Negara Malaysia, 1989).
Today, this statement remains highly relevant. For central banks, continuous research on the drivers of infl ation expectations, clarity in communication and educational pursuit in addressing inattention to infl ation – from the concept of infl ation to the technicalities of its formulation – will go a long way in preventing the unhinging of infl ation expectations.
References
Arioli, Rodolfo, Colm Bates, Heinz Dieden, Ioana Duca, Roberta Friz, Christian Gayer, Geoff Kenny, Aidan Meyler, Iskra Pavlova, “EU Consumers’ Quantitative Infl ation Perceptions and Expectations: An Evaluation,” European Central Bank Occasional Paper Series, 2017.
Bachmann, Rüdiger, Tim O. Berg, and Eric R. Sims, “Infl ation Expectations and Readiness to Spend: Cross-Sectional Evidence,” American Economic Journal: Economic Policy, 7(1), 1-35, 2015.
Bank Negara Malaysia, “Central Banking In An Era of Change, Landmark Speeches: 1959 –1988,” 1989.
Bank Negara Malaysia, “Evolving Household Balance Sheets and Implication for Private Consumption,” Annual Report, pages 70-73, 2013.
Baumeister, Roy F., Ellen Bratslavsky, Catrin Finkenauer, and Kathleen D. Vohs, “Bad is Stronger Than Good,” Review of General Psychology 5(4): 323-70, 2001.
Cavallo, Alberto, Guillermo Cruces, and Ricardo Perez-Truglia, “Infl ation Expectations, Learning, and Supermarket Prices: Evidence from Survey Experiments,” American Economic Journal: Macroeconomics 9(3): 1-35, 2017.
Coibion, Olivier, Yuriy Gorodnichenko, Saten Kumar, and Mathieu Pedemonte, “Infl ation Expectation as a Policy Tool?,” NBER Working Paper No. 24788, 2018.
Deutsche Bundesbank, “The Information Content of Survey Data on Expected Price Developments for Monetary Policy,” January Monthly Report 53(1), 35–49, 2001.
Doepke, Matthias, and Martin Schneider, “Infl ation and the Redistribution of Nominal Wealth,” Journal of Political Economy, Volume 114, (Number 6), 1069-109, 2006.
Ehrmann, Michael, Damjan Pfajar, and Emiliano Santoro, “Consumers’ Attitudes and Their Infl ation Expectations,” Finance and Economics Discussion Series, Divisions of Research & Statistics and Monetary Affairs, Federal Reserve Board, 2015.
Gamble, Amelie, “Euro Illusion or the Reverse? Effects of Currency and Income on Evaluation of Prices of Consumer Products,” Journal of Economics Psychology, 27(4), 531-540, 2006.
Gärling, Tommy, and Amelie Gamble, “Change in Perceived Value of Money without Change in Nominal Representation,” Göteborg Psychological Report No. 36:4, Göteborg University, Department of Psychology, 2006.
105Annual Report 2018
Outlook and
Policy in 2019
89Annual Report 2018
Gnan, Ernest, Johann Scharler, and Maria Antoinette Silgoner, “Infl ation Expectations – Role and Measurement for Monetary Policy,” Monetary Policy & the Economy, Oesterreichische Nationalbank (Austrian Central Bank), Issue 2, pages 41-67, 2009.
Hoover, Kevin D, “Phillips Curve,” The Concise Encyclopedia of Economics, 2008.
Ichiue, Hibiki, and Shusaku Nishiguchi, “Infl ation Expectations and Consumer Spending at the Zero Bound,” Bank of Japan Working Paper Series, 2013.
Kliesen, Kevin, “How Accurate Are Measures of Long-Term Infl ation Expectations?,” Federal Reserve Bank of St. Louis, 2015.
Leung, Christina, “The Demographics of Household Infl ation Perceptions and Expectations,” Reserve Bank of New Zealand, 2009.
Loke, Po Ling and Farina Adam Khong, “Infl ation: Perception vs. Reality,” Bank Negara Malaysia First Quarter 2017 Quarterly Bulletin, 2017.
Meyer, Brent, and Guhan Venkatu, “Demographic Differences in Infl ation Expectations: What Do They Really Mean?,” Federal Reserve Bank of Cleveland, 2011.
Mian, Atif, Rao Kamalesh, and Amir Sufi , “Household Balance Sheets, Consumption, and the Economic Slump,” The Quarterly Journal of Economics, Volume 128, Issue 4, Pages 1687–1726, 2013.
Mohanty, Deepak, “The Importance of Infl ation Expectations,” Speech at the S P Jain Institute of Management & Research, Mumbai, 2012.
Moore, Angus, “Measures of Infl ation Expectations in Australia,” Reserve Bank of Australia Bulletin, December 23-31, 2016.
Pasaogullari, Mehmet, and Timothy Bianco, “Survey-Based Measures of Infl ation Expectations,” Federal Reserve Bank of Cleveland, 2010.
Ranyard, Rob, Fabio Del Missier, Nicolao Bonini, Darren Duxbury, and Barbara Summers, “Perceptions and Expectations of Price Changes and Infl ation: A Review and Conceptual Framework,” Journal of Economic Psychology 29.4: 378-400, 2008.
Rowe, Jeremy, “How Are Households’ Infl ation Expectations Formed?,” Bank of England Quarterly Bulletin, vol. 56(2), pages 82-86, 2016.
Snir, Avichai and Daniel Levy, “Shrinking Goods and Sticky Prices: Theory and Evidence,” Working Paper Series 17/11, The Rimini Centre for Economic Analysis, 2011.
106 Annual Report 2018
Outlook and
Policy in 2019